Statistical Fragility of Steroid-Antiviral Therapy for Bell's Palsy: A Systematic Review of Randomized Controlled Trials.

Abstract

Oral corticosteroids are the current standard of care for Bell's Palsy (BP). Antivirals have also been incorporated as an adjunctive therapy, but randomized controlled trials (RCTs) evaluating the efficacy of combination antiviral-steroid therapy have yielded conflicting results. Here, we employ fragility index (FI) to evaluate the statistical reliability of RCTs investigating the clinical benefit of combination therapy. The underlying purpose of this study is to explore the robustness of using combinatorial treatment for BP. A systematic search of articles on PubMed, MEDLINE, and Embase from 1976 to 2024 identified relevant studies. A systematic review with statistical fragility analysis was conducted on RCTs that investigated steroid-antiviral versus steroid-only treatment for BP. FI, the number of outcome event reversals needed to flip the result of statistical significance, and reverse FI (rFI), the corollary to FI for nonsignificant outcomes, were calculated for all dichotomous variables. Eleven studies were included for analysis. Fifty-seven dichotomous outcome variables on recovery status were reported; mean FI was 5.561 (SD 4.778). Seven outcomes were statistically significant (P < .05) in favor of combinatorial treatment, with a mean FI of 2.714 (SD 2.984). Fifty outcomes were nonsignificant, with a mean rFI of 5.960 (SD 4.865). RCTs on steroid-antiviral treatment for BP show mixed effects with high statistical fragility, especially among statistically significant outcomes. Further studies are needed to increase the robustness of evidence for using combination therapy in BP to validate clinical decision-making.